A study of criteria for grading follicular lymphoma using a cell type classifier from pathology images based on complementary-label learning.

We propose a criterion for grading follicular lymphoma that is consistent with the intuitive evaluation, which is conducted by experienced pathologists. A criterion for grading follicular lymphoma is defined by the World Health Organization (WHO) based on the number of centroblasts and centrocytes within the field of view. However, the WHO criterion is not often used in clinical practice because it is impractical for pathologists to visually identify the cell type of each cell and count the number of centroblasts and centrocytes. Hence, based on the widespread use of digital pathology, we make it practical to identify and count the cell type by using image processing and then construct a criterion for grading based on the number of cells. Here, the problem is that labeling the cell type is not easy even for experienced pathologists. To alleviate this problem, we build a new dataset for cell type classification, which contains the pathologists' confusion records during labeling, and we construct the cell type classifier using complementary-label learning from this dataset. Then we propose a criterion based on the composition ratio of cell types that is consistent with the pathologists' grading. Our experiments demonstrate that the classifier can accurately identify cell types and the proposed criterion is more consistent with the pathologists' grading than the current WHO criterion.

Hydrodissection During Computed Tomography-Guided Cryoablation for Renal Tumors: Where is the Effective Fluid Accumulation Space in the Retroperitoneum?

PURPOSE: This study was performed to investigate the effectiveness of hydrodissection during computed tomography-guided renal cryoablation by evaluation of the fluid distribution based on the retroperitoneal anatomy with the interfascial plane. MATERIALS AND METHODS: Between March 2014 and March 2021, 52 renal tumors were treated by cryoablation with hydrodissection (36 men; mean age 72.5 years). The hydrodissection needle was located in perirenal space. The spreading fluid space based on the retroperitoneal anatomy with the interfascial plane was retrospectively evaluated. The fluid space that most effectively separated the tumor from the adjacent organs was defined. The relationship of the needle tip position in the perirenal space (renal capsule or fascia side) and the most effective fluid space was also evaluated. RESULTS: Cryoablation was successfully completed in all cases with no major complications. Hydrodissection was effective in all cases. The distance between the tumors and the adjacent organs was significantly longer after hydrodissection (from 7.50 ± 7.43 to 22.6 ± 9.86 mm) (P < 0.0001). Although fluid spreading through multiple retroperitoneal spaces was frequently observed, the retromesenteric plane was observed more frequently as the most effective fluid space (67.3%) than the perirenal space (21.2%) (P < 0.0001). Regardless of the needle tip position, the most effective fluid space was also commonly the retromesenteric plane. CONCLUSIONS: The retromesenteric plane could be the most effective fluid space to separate the tumor from the adjacent organ, regardless of where the hydrodissection needle tip is positioned in the perirenal space. LEVEL OF EVIDENCE: 3b.

Subtype classification of malignant lymphoma using immunohistochemical staining pattern.

PURPOSE: For the image classification problem, the construction of appropriate training data is important for improving the generalization ability of the classifier in particular when the size of the training data is small. We propose a method that quantitatively evaluates the typicality of a hematoxylin-and-eosin (H&E)-stained tissue slide from a set of immunohistochemical (IHC) stains and applies the typicality to instance selection for the construction of classifiers that predict the subtype of malignant lymphoma to improve the generalization ability. METHODS: We define the typicality of the H&E-stained tissue slides by the ratio of the probability density of the IHC staining patterns on low-dimensional embedded space. Employing a multiple-instance-learning-based convolutional neural network for the construction of the subtype classifier without the annotations indicating cancerous regions in whole slide images, we select the training data by referring to the evaluated typicality to improve the generalization ability. We demonstrate the effectiveness of the instance selection based on the proposed typicality in a three-class subtype classification of 262 malignant lymphoma cases. RESULTS: In the experiment, we confirmed that the subtypes of typical instances could be predicted more accurately than those of atypical instances. Furthermore, it was confirmed that instance selection for the training data based on the proposed typicality improved the generalization ability of the classifier, wherein the classification accuracy was improved from 0.664 to 0.683 compared with the baseline method when the training data was constructed focusing on typical instances. CONCLUSION: The experimental results showed that the typicality of the H&E-stained tissue slides computed from IHC staining patterns is useful as a criterion for instance selection to enhance the generalization ability, and this typicality could be employed for instance selection under some practical limitations.

Manifold Modeling in Embedded Space: An Interpretable Alternative to Deep Image Prior.

Deep image prior (DIP), which uses a deep convolutional network (ConvNet) structure as an image prior, has attracted wide attention in computer vision and machine learning. DIP empirically shows the effectiveness of the ConvNet structures for various image restoration applications. However, why the DIP works so well is still unknown. In addition, the reason why the convolution operation is useful in image reconstruction, or image enhancement is not very clear. This study tackles this ambiguity of ConvNet/DIP by proposing an interpretable approach that divides the convolution into "delay embedding" and "transformation" (i.e., encoder-decoder). Our approach is a simple, but essential, image/tensor modeling method that is closely related to self-similarity. The proposed method is called manifold modeling in embedded space (MMES) since it is implemented using a denoising autoencoder in combination with a multiway delay-embedding transform. In spite of its simplicity, MMES can obtain quite similar results to DIP on image/tensor completion, super-resolution, deconvolution, and denoising. In addition, MMES is proven to be competitive with DIP, as shown in our experiments. These results can also facilitate interpretation/characterization of DIP from the perspective of a "low-dimensional patch-manifold prior."

Real-time estimation of electric fields induced by transcranial magnetic stimulation with deep neural networks.

BACKGROUND: Transcranial magnetic stimulation (TMS) plays an important role in treatment of mental and neurological illnesses, and neurosurgery. However, it is difficult to target specific brain regions accurately because the complex anatomy of the brain substantially affects the shape and strength of the electric fields induced by the TMS coil. A volume conductor model can be used for determining the accurate electric fields; however, the construction of subject-specific anatomical head structures is time-consuming. OBJECTIVE: The aim of this study is to propose a method to estimate electric fields induced by TMS from only T1 magnetic resonance (MR) images, without constructing a subject-specific anatomical model. METHODS: Very large sets of electric fields in the brain of subject-specific anatomical models, which are constructed from T1 and T2 MR images, are computed by a volume conductor model. The relation between electric field distribution and T1 MR images is used for machine learning. Deep neural network (DNN) models are applied for the first time to electric field estimation. RESULTS: By determining the relationships between the T1 MR images and electric fields by DNN models, the process of electric field estimation is markedly accelerated (to 0.03 s) due to the absence of a requirement for anatomical head structure reconstruction and volume conductor computation. Validation shows promising estimation accuracy, and rapid computations of the DNN model are apt for practical applications. CONCLUSION: The study showed that the DNN model can estimate the electric fields from only T1 MR images and requires low computation time, suggesting the possibility of using machine learning for real-time electric field estimation in navigated TMS.

Robust 3D image reconstruction of pancreatic cancer tumors from histopathological images with different stains and its quantitative performance evaluation.

PURPOSE: Histopathological imaging is widely used for the analysis and diagnosis of multiple diseases. Several methods have been proposed for the 3D reconstruction of pathological images, captured from thin sections of a given specimen, which get nonlinearly deformed due to the preparation process. The majority of the available methods for registering such images use the degree of matching of adjacent images as the criteria for registration, which can result in unnatural deformations of the anatomical structures. Moreover, most methods assume that the same staining is used for all images, when in fact multiple staining is usually applied in order to enhance different structures in the images. METHODS: This paper proposes a non-rigid 3D reconstruction method based on the assumption that internal structures on the original tissue must be smooth and continuous. Landmarks are detected along anatomical structures using template matching based on normalized cross-correlation (NCC), forming jagged shape trajectories that traverse several slices. The registration process smooths out these trajectories and deforms the images accordingly. Artifacts are automatically handled by using the confidence of the NCC in order to reject unreliable landmarks. RESULTS: The proposed method was applied to a large series of histological sections from the pancreas of a KPC mouse. Some portions were dyed primarily with HE stain, while others were dyed alternately with HE, CK19, MT and Ki67 stains. A new evaluation method is proposed to quantitatively evaluate the smoothness and isotropy of the obtained reconstructions, both for single and multiple staining. CONCLUSIONS: The experimental results show that the proposed method produces smooth and nearly isotropic 3D reconstructions of pathological images with either single or multiple stains. From these reconstructions, microanatomical structures enhanced by different stains can be simultaneously observed.

Optimization of the Photon Path Length Probability Density Function-Simultaneous (PPDF-S) Method and Evaluation of CO2 Retrieval Performance Under Dense Aerosol Conditions.

The photon path length probability density function-simultaneous (PPDF-S) algorithm is effective for retrieving column-averaged concentrations of carbon dioxide (XCO2) and methane (XCH4) from Greenhouse gases Observing Satellite (GOSAT) spectra in Short Wavelength InfraRed (SWIR). Using this method, light-path modification attributable to light reflection/scattering by atmospheric clouds/aerosols is represented by the modification of atmospheric transmittance according to PPDF parameters. We optimized PPDF parameters for a more accurate XCO2 retrieval under aerosol dense conditions based on simulation studies for various aerosol types and surface albedos. We found a more appropriate value of PPDF parameters referring to the vertical profile of CO2 concentration as a measure of a stable solution. The results show that the constraint condition of a PPDF parameter that represents the light reflectance effect by aerosols is sufficiently weak to affect XCO2 adversely. By optimizing the constraint, it was possible to obtain a stable solution of XCO2. The new optimization was applied to retrieval analysis of the GOSAT data measured in Western Siberia. First, we assumed clear sky conditions and retrieved XCO2 from GOSAT data obtained near Yekaterinburg in the target area. The retrieved XCO2 was validated through a comparison with ground-based Fourier Transform Spectrometer (FTS) measurements made at the Yekaterinburg observation site. The validation results showed that the retrieval accuracy was reasonable. Next, we applied the optimized method to dense aerosol conditions when biomass burning was active. The results demonstrated that optimization enabled retrieval, even under smoky conditions, and that the total number of retrieved data increased by about 70%. Furthermore, the results of the simulation studies and the GOSAT data analysis suggest that atmospheric aerosol types that affected CO2 analysis are identifiable by the PPDF parameter value. We expect that we will be able to suggest a further improved algorithm after the atmospheric aerosol types are identified.

Super-Resolution of Magnetic Resonance Images via Convex Optimization with Local and Global Prior Regularization and Spectrum Fitting.

Given a low-resolution image, there are many challenges to obtain a super-resolved, high-resolution image. Many of those approaches try to simultaneously upsample and deblur an image in signal domain. However, the nature of the super-resolution is to restore high-frequency components in frequency domain rather than upsampling in signal domain. In that sense, there is a close relationship between super-resolution of an image and extrapolation of the spectrum. In this study, we propose a novel framework for super-resolution, where the high-frequency components are theoretically restored with respect to the frequency fidelities. This framework helps to introduce multiple simultaneous regularizers in both signal and frequency domains. Furthermore, we propose a new super-resolution model where frequency fidelity, low-rank (LR) prior, low total variation (TV) prior, and boundary prior are considered at once. The proposed method is formulated as a convex optimization problem which can be solved by the alternating direction method of multipliers. The proposed method is the generalized form of the multiple super-resolution methods such as TV super-resolution, LR and TV super-resolution, and the Gerchberg method. Experimental results show the utility of the proposed method comparing with some existing methods using both simulational and practical images.

Semi-Automated Biomarker Discovery from Pharmacodynamic Effects on EEG in ADHD Rodent Models.

We propose a novel semi-automatic approach to design biomarkers for capturing pharmacodynamic effects induced by pharmacological agents on the spectral power of electroencephalography (EEG) recordings. We apply this methodology to investigate the pharmacodynamic effects of methylphenidate (MPH) and atomoxetine (ATX) on attention deficit/hyperactivity disorder (ADHD), using rodent models. We inject the two agents into the spontaneously hypertensive rat (SHR) model of ADHD, the Wistar-Kyoto rat (WKY), and the Wistar rat (WIS), and record their EEG patterns. To assess individual EEG patterns quantitatively, we use an integrated methodological approach, which consists of calculating the mean, slope and intercept parameters of temporal records of EEG spectral power using a smoothing filter, outlier truncation, and linear regression. We apply Fisher discriminant analysis (FDA) to identify dominant discriminants to be heuristically consolidated into several new composite biomarkers. Results of the analysis of variance (ANOVA) and t-test show benefits in pharmacodynamic parameters, especially the slope parameter. Composite biomarker evaluation confirms their validity for genetic model stratification and the effects of the pharmacological agents used. The methodology proposed is of generic use as an approach to investigating thoroughly the dynamics of the EEG spectral power.

Enhanced Methane Emissions during Amazonian Drought by Biomass Burning.

The Amazon is a significant source of atmospheric methane, but little is known about the source response to increasing drought severity and frequency. We investigated satellite observations of atmospheric column-averaged methane for the 2010 drought and subsequent 2011 wet year in the Amazon using an atmospheric inversion scheme. Our analysis indicates an increase in atmospheric methane over the southern Amazon region during the drought, representing an increase in annual emissions relative to the wet year. We attribute the increase to emissions from biomass burning driven by intense drought, combined with carbon monoxide showing seasonal variations corresponding to methane variations. We show that there is probably a strong correspondence between drought and methane emissions in the Amazon.

Simultaneous retrieval of atmospheric CO2 and light path modification from space-based spectroscopic observations of greenhouse gases: methodology and application to GOSAT measurements over TCCON sites.

This paper presents an improved photon path length probability density function method that permits simultaneous retrievals of column-average greenhouse gas mole fractions and light path modifications through the atmosphere when processing high-resolution radiance spectra acquired from space. We primarily describe the methodology and retrieval setup and then apply them to the processing of spectra measured by the Greenhouse gases Observing SATellite (GOSAT). We have demonstrated substantial improvements of the data processing with simultaneous carbon dioxide and light path retrievals and reasonable agreement of the satellite-based retrievals against ground-based Fourier transform spectrometer measurements provided by the Total Carbon Column Observing Network (TCCON).

A quadratically constrained MAP classifier using the mixture of Gaussians models as a weight function.

In this paper, we propose classifiers derived from quadratically constrained maximum a posteriori (QCMAP) estimation. The QCMAP consists of the maximization of the expectation of a cost function, which is derived from the maximum a posteriori probability and a quadratic constraint. This criterion is highly general since its forms include least squares regressions and a support vector machine. Furthermore, the criterion provides a novel classifier, the "Gaussian QCMAP." The QCMAP procedure still has large theoretical interest and its full extensibility has yet to be explored. In this paper, we propose using the mixture of Gaussian distributions as the QCMAP weight function. The mixture of Gaussian distributions has wide-ranging applicability, and encompasses forms, such as a normal distribution model and a kernel density model. We propose four types of mixture of Gaussian functions for QCMAP classifiers, and conduct experiments to demonstrate their advantages.

Mountaintop observation of CO2 absorption spectra using a short wavelength infrared Fourier transform spectrometer.

The absorption spectra of surface-scattered solar radiation were measured from the top of Mount Tsukuba (altitude 833 m) in the short wavelength infrared region using a Fourier transform spectrometer (FTS). The FTS used in this experiment was the breadboard model of the FTS on the Greenhouse Gases Observing Satellite, which was launched on 23 January 2009. In situ measurement of carbon dioxide (CO(2)) from a Cessna airplane was performed simultaneously with the FTS observation. The CO(2) column abundances were retrieved from the observed spectra under the assumption of the absence of aerosol. The retrieved CO(2) column abundances over a few minutes dispersed within 1%. The remaining bias was considered to be caused by the no-aerosol assumption.

Interpolated local model fitting method for accurate and fast single-shot surface profiling.

The local model fitting (LMF) method is a useful single-shot surface profiling algorithm based on spatial carrier frequency fringe patterns. The measurement principle of the LMF method relies on the assumption that the target surface is locally flat. In this paper, we first analyze the measurement error of the LMF method caused by violation of the locally flat assumption. More specifically, we theoretically prove that the measurement error is zero at fringe intensity extrema in an interference pattern even when the locally flat assumption is violated. Based on this theoretical finding, we propose a new surface profiling method called the interpolated LMF (iLMF) algorithm, which is more accurate and computationally efficient than the original LMF method. The practical usefulness of the iLMF method is shown through experiments.

Retrieval of atmospheric methane from high spectral resolution satellite measurements: a correction for cirrus cloud effects.

We assessed the accuracy of methane (CH(4)) retrievals from synthetic radiance spectra particular to Greenhouse Gases Observing Satellite observations. We focused on estimating the CH(4) vertical column amount from an atmosphere that includes thin cirrus clouds, taking into account uncertain meteorological conditions. A photon path-length probability density function (PPDF)-based method was adapted to correct for atmospheric scattering effects in CH(4) retrievals. This method was shown to provide similar retrieval accuracy as compared to a carbon dioxide (CO(2))-proxy-based correction approach. It infers some advantages of PPDF-based method for methane retrievals under high variability of CO(2) abundance.

Tangent height registration method for the Version 1.4 data retrieval algorithm of the solar occultation sensor ILAS-II.

The Improved Limb Atmospheric Spectrometer-II (ILAS-II) is a satellite-borne solar occultation sensor onboard the Advanced Earth Observing Satellite-II (ADEOS-II). The ILAS-II succeeded the ILAS. The ILAS-II used four grating spectrometers to observe vertical profiles of gas volume mixing ratios of trace constituents and was also equipped with a Sun-edge sensor to determine tangent heights geometrically with high precision. The accuracy of gas volume mixing ratios depends on the accuracy of the tangent height determination. The combination method is a tangent height registration method that was developed to give appropriate tangent heights for the ILAS-II Version 1.4 data retrieval algorithm. This study describes the method used in the ILAS-II Version 1.4 retrieval algorithm to register tangent heights. The root-sum-square total random error is estimated to be 30 m, and the total systematic error is 180 m at an altitude of 30 km. The influence of the tangent height errors on the vertical profiles of gas volume mixing ratios in ILAS-II Version 1.4 is estimated by using the relative difference. The relative difference for each species is within 7% (20%) for an altitude shift of +/-100 m(+/-300 m).

Parameterization of aerosol and cirrus cloud effects on reflected sunlight spectra measured from space: application of the equivalence theorem.

An original methodology to account for aerosol and cirrus cloud contributions to reflected sunlight is described. This method can be applied to the problem of retrieving greenhouse gases from satellite-observed data and is based on the equivalence theorem with further parameterization of the photon path-length probability density function (PPDF). Monte Carlo simulation was used to validate this parameterization for a vertically nonhomogeneous atmosphere including an aerosol layer and cirrus clouds. Initial approximation suggests that the PPDF depends on four parameters that can be interpreted as the effective cloud height, cloud relative reflectance, and two additional factors to account for photon path-length distribution under the cloud. We demonstrate that these parameters can be efficiently retrieved from the nadir radiance measured in the oxygen A-band and from the H(2)O-saturated area of the CO(2) 2.0 microm spectral band.

[A case of primary Sjögren' s syndrome with pathological findings resembling multicentric Castleman' s disease showed multiple nodular opacities on chest CT scans].

A 49-year-old woman with primary Sjögren's syndrome a few years previously was admitted to our hospital complaining of tongue and skin eruptions, swelling of the face and neck and for examination of liver injury and hypereosinophilia. A blood test revealed leukocytosis with eosinophilia, mild liver injury, polyclonal hypergammaglobulinemia, and positive results for anti-nuclear antibody, anti-SS-A antibody and anti-SS-B antibody. Chest CT scan showed multiple nodular opacities with cavities in peripheral regions of both lungs. Biopsy specimens from the right lower lobe obtained by video-assisted thoracoscopy revealed marked infiltration of plasma cells and lymphocytes in alveolar lumina, lymph follicles with germinal centers in other areas of the pulmonary parenchyma, and lymphocytes infiltrate in alveolar wall adjacent bronchi and bronchioles. The histological diagnosis was pulmonary involvement of multicentric Castleman's disease. This was a rare case of Sjögren's syndrome accompanied by pathological findings of multicentric Castleman' s disease showed multiple nodular opacities in chest CT scans.

Simultaneous stratospheric gas and aerosol retrievals from broadband infrared occultation measurements.

The inversion method for simultaneous gas (O3, NO2, HNO3, N2O, CH4, H2O, CFC-11, CFC-12, N2O5, and ClONO2) and aerosol retrievals from broadband continuous IR spectra of occultation measurements is described. Both gas and aerosol physical modeling with consideration of the multicomponent character of aerosol and polar stratospheric clouds (PSCs) are used to minimize the difference between measured and modeled transmittance spectra under smoothness constraints imposed on particle-size distributions for each PSC component and positive constraints on all gas and aerosol parameters. The method is tested by numerical simulations in which synthetic occultation measurements inherent to the improved limb atmospheric spectrometer are used. The study reveals that the method has significant advantages over other approaches based on offset or gas-window-channel aerosol correction for accurate gas retrievals and provides additional information on the particle-size composition, volume density, and chemical component character of PSCs.

Retrieval of trace gases from aerosol-influenced infrared transmission spectra observed by low-spectral-resolution Fourier-transform spectrometers.

A method for the simultaneous retrieval of gas concentrations and an extinction spectrum of aerosols and polar stratospheric clouds from infrared transmission spectra observed in the solar occultation geometry is described. It is particularly suited to measurements by Fourier-transform spectrometers with relatively low spectral resolution (0.1-1 cm(-1)). The method does not require a priori assumptions on aerosol properties; it utilizes only the fact that the wave-number dependence of aerosol extinction is much weaker than that of gas absorption. In this method, an aerosol extinction spectrum is approximated by a straight line within a relatively wide spectral range defined as mediumwindow.